基于效用驱动的网格资源协同预留策略

doi:10.3969/j.issn.1000-436x.2014.05.014

通信学报 ›› 2014, Vol. 35 ›› Issue (5): 101-107.doi: 10.3969/j.issn.1000-436x.2014.05.014

基于效用驱动的网格资源协同预留策略

丁长松^1,²,王志英²,胡志刚³

¹ 湖南中医药大学管理与信息工程学院，湖南长沙 410208
² 国防科技大学计算机学院，湖南长沙 410073
³ 中南大学软件学院，湖南长沙 410083

出版日期:2014-05-25 发布日期:2017-07-24
基金资助:
国家自然科学基金资助项目;博士点基金资助项目;湖南省科技计划基金资助项目;湖南省科技计划基金资助项目;湖南省教育厅科学研究优秀青年基金资助项目

Utility-driven based co-allocation resource reservation strategy in computational grid

Chang-song DING^1,²,Zhi-ying WANG²,Zhi-gang HU³

¹ School of Administration and Information Engineering, Hunan University of Chinese Medicine, Changsha 410208, China
² College of Computer, National University of Defense Technology, Changsha 410073, China
³ School of Software, Central South University, Changsha 410083, China

Online:2014-05-25 Published:2017-07-24
Supported by:
The National Natural Science Foundation of China;The Doctoral Program of Higher Education of China;The Science and Technology Projects Fund of Hunan Province;The Science and Technology Projects Fund of Hunan Province;Scientific Research Fund for Outstanding Young Teachers of Hunan Provincial Education Department

摘要/Abstract

摘要：

针对资源预留过程中价格对市场竞争力的影响导致收益不确定性问题，提出一种可量化分析价格、资源竞争力以及收益三者关系的协同预留策略。该策略基于本地任务的真实相关统计特性，在有效保障网格任务QoS与本地任务QoS基础上，通过价格调整来平衡资源提供方的市场竞争力与收益之间的冲突。理论分析给出了预留策略的有效性证明和预留算法，仿真实验采用真实网格系统中任务负载信息作为实验负载，并在模拟网格系统中对预留策略的性能表现进行了检验。实验结果表明，该策略在均衡资源负载、平衡资源节点相对收益率以及保障任务QoS方面的性能表现显著优于传统的预留策略。

关键词: 网格计算, 效用驱动, 资源预留, 定价策略

Abstract:

Profit of resource provider is always uncertain because price of resource has important impact on market competi-tiveness. A co-reservation strategy was presented, which could be used to quantitative analysis the relationship of reservation price, resource competitiveness and profit. Based on real statistical characteristic of local tasks, the model provids the grid job QoS guarantee and local job QoS guarantee, which balances the conflict between market competitiveness and profit of resource provider by efficient price adjustment. The validity of the model and its algorithm were presented theoreti-cally. The performance of the proposed strategy was simulated in a grid simulation system using the real task load of the practical grid system. The results show that the proposed co-reservation strategy outperforms traditional reservation strategy in terms of balancing resource load, profit rate of resource node and QoS guarantee.

Key words: grid computing, utility-driven, resource reservation, pricing strategy

丁长松,王志英,胡志刚. 基于效用驱动的网格资源协同预留策略[J]. 通信学报, 2014, 35(5): 101-107.

Chang-song DING,Zhi-ying WANG,Zhi-gang HU. Utility-driven based co-allocation resource reservation strategy in computational grid[J]. Journal on Communications, 2014, 35(5): 101-107.

图/表 3

参考文献 25

[1]	邓水光，黄龙涛，吴斌等．一种QoS最优的语义Web服务自动组合方法. 计算机学报， 2013,36(5): 1015-1030. DENG S G , HUANG L T , WU B , et al. QoS optinmal automatic com-position of semantic Web services[J]. Chinese Journal of Com-puters, 2013,36(5): 1015-1030.
[2]	FOSTER I . What is the grid? a three point checklist[J]. GRID Today, 2002,1(6):20-27.
[3]	TAKEFUSA A , NAKADA H , KUDOH T , et al. An advance reserva-tion-based co-allocation algorithm for distributed computers and net-work bandwidth on QoS guaranteed grids[A]. Job Scheduling Strate-gies for Parallel Processing, Ser Lecture Notes in Computer Sci-ence[C]. Springer Berlin/Heidelberg, 2010. 16-34.
[4]	MAKHLOUF S A , YAGOUBI B . Resources co-allocation strategies in grid computing[A]. Proceedings of the Third International Conference on Computer Science and its Applications (CIIA'11)[C]. Saida, Algeria, 2011.
[5]	MAKHLOUF E , TORDSSON J . A grid resource broker supporting ad-vance reservations and benchmark-based resource selection[A]. Lec-ture Notes in Computer Science[C]. Springer-Verlag, 2005.1077-1085.
[6]	丁长松，胡志刚，胡周君．基于收益与风险均衡的网格资源预留策略[J]．华南理工大学学报(自然科学版), 2011,39(2): 88-94. DING C S , HU Z G , HU Z J . Grid resource reservation policy based on the balance of benefits and risks[J]. Journal of South China Univer-sity of Technology(Natural Science Edition), 2011,39(2): 88-94.
[7]	FOSTER I , ROY A , SANDER V , et al. A Quality of Service Architec-ture that Combines Resource Reservation and Application Adapta-tion[R]. 2000.
[8]	SODAN A C , DOSHI C , BARSANTI L , et al. scheduling and adaptive resource allocation to mitigate advance reservation impact[A]. IEEE International Symposium on Cluster Computing and the Grid[C]. 2006. 1-8.
[9]	WU M , SUN X H , CHEN Y . QoS oriented resource reservation in shared environments[A]. Proc of the 6th IEEE International Sympo-sium on Cluster Computing and the Grid[C]. 2006. 601-608.
[10]	LEE Y C , WANG C , ZOMAYA A Y , et al. Profit-driven scheduling for cloud services with data access awareness[J]. Journal of Parallel and Distributed Computing, 2012,72(4): 591-602.
[11]	SNELL Q , CLEMENT M , JACKSONA D , et al. The performance impact of advance reservation meta-scheduling[J]. Lecture Notes in Computer Science, 2000,1911: 137-153.
[12]	SABITHA B S , RANI R , VENKATESAN R R . Resource reservation in grid computing environments: design issues[A]. IEEE 2011 3rd In-ternational Conference on Electronics Computer Technology (ICECT)[C]. 2011. 66-70.
[13]	XING J B , WU C , TAO M L , et al. Flexible advance reservation for grid computingc[A]. The 2nd International Workshop on Grid and Cooperative Computing(GCC'03)[C]. Shaihai, China, 2003.
[14]	KAUSHIK N R , FIGUEIRA S M , CHIAPPARI S A . Flexible time-windows for advance reservation scheduling[A]. Proceedings of IEEE International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems[C]. California, 2006.218-225.
[15]	TOMáS L , CAMINERO B , CARRIóN C . Improving grid resource usage: metrics for measuring fragmentation[A]. Proceedings of the 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing[C]. IEEE Computer Society, 2012.352-359.
[16]	TOMáS L , CAMINERO B , CARRIóN C , et al. On the improvement of grid resource utilization: preventive and reactive rescheduling ap-proaches[A]. Journal of Grid Computing, 2012,10(3): 475-499.
[17]	TOMáS L , CAMINERO B , CARRIóN C . Bag of tasks rescheduling within real grid environments: different approaches[A]. 21st Euromi-cro International Conference on IEEE Parallel, Distributed and Net-work-Based Processing (PDP)[C]. 2013. 213-217.
[18]	CAO J W , ZIMMERMANN F . Queue scheduling and advance reser-vations with COSY[A]. Proceedings of International Symposium on Parallel and Distributed Processing[C]. Santa Fe, New Mexico, 2004.
[19]	BUYYA R , ABRAMSON D , VENUGOPAL S . The grid economy[J]. Proceedings of the IEEE, 2005,93(3): 698-714.
[20]	SULISTIO A , KIM K H , BUYYA R . Using revenue management to determine pricing of reservations[A]. Third IEEE International Con-ference on e-Science and Grid Computing[C]. 2007. 396-404.
[21]	SINGH G , KESSELMAN C , DEELMAN E . Adaptive pricing for resource reservations in shared environments[A]. The 8th Grid Com-puting Conference[C]. 2007. 74-80.
[22]	LUBLIN U , FEITELSON D G . The workload on parallel supercom-puters: modeling the characteristics of rigid jobs[J]. Journal of Parallel and Distributed Computing, 2003,63(11): 1105-1122.
[23]	TSAKALOZOS K , KLLAPI H , SITARIDI E , et al. Flexible use of cloud resources through profit maximization and price discrimina-tion[J]. IEEE 27th International Conference on Data Engineering (ICDE)[C]. 2011. 75-86.
[24]	BAL H E , BHOEDJANG R R , HOFMAN R , et al. The distributed ASCI supercomputer project[J]. ACM Operating Systems Review, 2000,34(4): 76-96.
[25]	BUYYA R . Tconomic-based Distributed Resource Management and Scheduling for Grid Computing[D]. Australia: Monash University, 2002.

模型	接纳任务数	违约率	资源利用率	系统总收益
UD	9326	5.61%	64.38%	8.64×10⁶
SPA	6784	13.21%	35.68%	4.47×10⁶
RTR	7654	8.94%	57.32%	4.96×10⁶

策略指标	任务数
策略指标	2000	4000	6000	8000	10000
MUR/%	37.23	71.37	87.54	71.79	53.97
MP	27.32	46.26	51.39	55.46	54.27
MPB/%	19.10	25.36	10.23	5.16	3.72

基于效用驱动的网格资源协同预留策略

Utility-driven based co-allocation resource reservation strategy in computational grid

在线阅读

PDF下载

可视化

被引次数

摘要/Abstract

引用本文

使用本文

图/表 3

参考文献 25

相关文章 15

Metrics

推荐阅读 0

[1]	孙国林,欧睿杰,刘贵松. 基于深度强化学习的应急物联网切片资源预留算法[J]. 通信学报, 2020, 41(9): 8-20.
[2]	丛鑫,訾玲玲,孙劲光. 基于可信拍卖机制的视频移植定价策略[J]. 通信学报, 2016, 37(4): 1-65.
[3]	丁长松1,2，王志英2，胡志刚3. 基于效用驱动的网格资源协同预留策略[J]. 通信学报, 2014, 35(5): 14-107.
[4]	丛鑫,双错,苏森,杨放春,訾玲玲. 基于带宽请求预测和云资源预留的视频移植策略[J]. 通信学报, 2014, 35(5): 167-174.
[5]	丛鑫1,2，双锴1，苏森1，杨放春1，訾玲玲2. 基于带宽请求预测和云资源预留的视频移植策略[J]. 通信学报, 2014, 35(5): 22-175.
[6]	沈时军1，刘欣然1,2，张鸿1，朱春鸽1,2. 云计算中的服务可用性保障机制[J]. 通信学报, 2014, 35(2): 25-206.
[7]	沈时军,刘欣然,张鸿,朱春鸽. 云计算中的服务可用性保障机制[J]. 通信学报, 2014, 35(2): 202-206.
[8]	张永晖,林漳希,刘建华,梁泉. 基于AAA认证的仓储移动网络安全关联转移方案[J]. 通信学报, 2012, 33(Z1): 186-191.
[9]	肖鹏,胡志刚,阎朝坤,屈喜龙. 面向网格工作流的松弛预留策略[J]. 通信学报, 2012, 33(2): 23-29.
[10]	丁长松,胡志刚,肖鹏. 网格环境中可靠性增强的资源预留策略[J]. 通信学报, 2011, 32(7): 40-46.
[11]	熊曾刚,杨扬,曾明. 基于Petri网的两阶段网格任务调度模型与分析[J]. 通信学报, 2009, 30(8): 69-77.
[12]	王贤茜,王淼,陆月明,纪越峰. 串行同源故障通告机制及其在ASON中的应用[J]. 通信学报, 2009, 30(5): 82-87.
[13]	邓勇,陈建刚,王汝传,张琳. 网格计算环境的一种基于信任度的授权委托机制[J]. 通信学报, 2008, 29(9): 10-17.
[14]	舒万能. 网格计算中资源负载均衡问题的优化调度模型及其算法[J]. 通信学报, 2008, 29(3A): 50-55.
[15]	季一木,王汝传. 基于粒子群的网格任务调度算法研究[J]. 通信学报, 2007, 28(10): 60-66.